JPH08138711A - Fuel cell driving device, catalyst poisoned ratio detecting device, and fuel cell system - Google Patents

Abstract

PURPOSE: To prevent control delay and stall in case of drop in cell output caused by a catalyst being poisoned by arranging a plurality of specified means in a driving device of a fuel cell producing electromotive force by a specified chemical reaction. CONSTITUTION: A driving device of a fuel cell generating electromotive force by the chemical reaction of a reaction gas (manufactured in a reformer 16) in an electrode (in a cell stack 10) on which a catalyst is carried has an initial cell performance memorizing device (A) 84, a poisoned state detecting means (B) 30, an output signal detecting means 32 for detecting an output signal of the cell stack 10 when the detecting means (B) detects that the catalyst is poisoned, and a cell performance operating means (CPU: 82, RAM: 86, ROM: 84) for calculating a ratio corresponding to poisoned ratio of the catalyst from the output signal detected and cell performance memorized by the memorizing device (A) and for finding cell performance in the poisoned state of the cell from the ratio calculated and the cell performance.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a fuel cell driving apparatus for supplying a reaction gas to an electrode carrying a catalyst to obtain an electromotive force from a chemical reaction of the reaction gas, and a poisoning of the catalyst of the fuel cell. A fuel cell catalyst poisoning rate detection device for detecting the rate,
Fuel cell system and.

[0002]

2. Description of the Related Art Generally, a fuel cell is known as a device for directly converting the energy of fuel into electrical energy. In a fuel cell, usually, a pair of electrodes are arranged with an electrolyte sandwiched between them, a reaction gas of hydrogen (fuel gas) is brought into contact with the surface of one electrode, and an oxidizing gas containing oxygen is also supplied to the surface of the other electrode. The electrodes are brought into contact with each other, and the electrochemical reaction that occurs at this time is used to extract electrical energy from between the electrodes.

In such a fuel cell, the characteristics of the cell are determined by the current-voltage characteristics, but the cell characteristics change depending on various driving conditions (factors that can be directly controlled).
For example, the gas pressure, the gas utilization rate, the battery temperature, and the like correspond to the driving conditions, but the battery output may decrease even if these driving conditions are constant. This is because the battery characteristics also change due to factors that cannot be directly controlled.
The battery output also decreases due to the electrode being too wet, the electrode being too dry, and carbon monoxide poisoning of the catalyst. When these factors are added, the battery output decreases and it is not known what the battery characteristics will be.

Under such a background, conventionally, when the fuel cell falls below a predetermined current-voltage characteristic, the driving condition is controlled so that the fuel cell is operated under a normal current-voltage.
A drive device for a fuel cell that adjusts to voltage characteristics has been proposed (paragraph 0012 of JP-A-5-47394). When the electrodes get too wet, the cell output begins to gradually decrease until it reaches an output below the predetermined current-voltage characteristic.However, according to this drive device, dry air can be sent to the fuel cell cells. As a result, the electrode was overwetted and gradually dried, and as a result, normal current-voltage characteristics could be obtained.

[0005]

However, in this conventional fuel cell driving apparatus, since the fuel cell causes a correction operation after the fuel cell has an output below the predetermined current-voltage characteristic, a normal correction operation is performed. It took some time to return to a good current-voltage characteristic. For this reason, control delay occurred in the battery output.

Further, the correction operation is not performed with the certainty that the target current-voltage characteristics will be restored if the control target is changed from the beginning, but the correction operation is gradually changed. When the target is reached, the control is stopped there. In other words, it is a fumbling control, and this also makes the aforementioned control delay large.

Further, if the amount of change that is changed little by little is too large, the control may overshoot, which may cause the fuel cell main body to stall (the output of the fuel cell may drastically decrease).

Such control delay and stall of the battery output can occur even when the catalyst is poisoned by carbon monoxide. The drive device for a fuel cell according to the present invention has been made in view of these problems, and an object thereof is to eliminate a control delay or a stall when the cell output is reduced due to poisoning of a catalyst of the fuel cell. I am trying.

On the other hand, the catalyst poisoning rate detecting device for a fuel cell of the present invention is intended to quantitatively detect poisoning by carbon monoxide in a catalyst. Conventionally, there has been a configuration that indirectly detects how poisoned a catalyst is by detecting the concentration of carbon monoxide in the fuel gas, but with this configuration, how many parts of the catalyst are poisoned. I couldn't ask what was happening. That is, conventionally, there is no method for quantitatively determining the poisoning of the catalyst, and the present invention aims to enable the quantitative detection.

[0010]

In order to achieve such an object, the following constitution is adopted as a means for solving the above problems.

That is, the fuel cell driving apparatus of the present invention is a fuel cell driving apparatus in which a reaction gas is supplied to an electrode carrying a catalyst to obtain electromotive force from a chemical reaction of the reaction gas. Initial battery characteristic storage means for storing the battery characteristics in the initial state in advance, poisoning state detection means for detecting that the catalyst is in a poisoned state, and the catalyst is poisoned state by the poisoned state detection means. When it is detected that the output signal of the fuel cell is detected, the output signal of the fuel cell detected and the cell characteristics stored by the initial cell characteristic storage means The gist is that a ratio is calculated according to the poisoning rate of the catalyst, and a battery characteristic calculation means for calculating the battery characteristic in the poisoned state of the fuel cell from the ratio and the cell characteristic is provided. Claim 1 as described).

Here, the cell characteristics are current-voltage characteristics showing the relationship between the current and voltage of the fuel cell. Further, the initial state of the fuel cell means a steady state before catalyst poisoning, which causes a reduction in cell output, occurs.In other words, the supplied fuel gas does not contain any carbon monoxide. It means a state containing only hydrogen. The poisoning rate of the catalyst represents the degree of poisoning of the catalyst by carbon monoxide.

In the fuel cell driving device having the above structure,
It is preferable to have a configuration including operation permission region setting means for setting an operation region in which the operation of the fuel cell is permitted from various conditions that limit the operation of the fuel cell (claim 2).
Those listed).

In the fuel cell drive device,
The battery characteristics at the time of the poisoning state calculated by the battery characteristic calculation means is limited to the operation region set by the operation permission region setting means, and on the limited battery characteristics,
Determination means for determining whether or not the control operating point can be moved,
It may be configured to include a control system switching permission unit that permits switching of a control system of operation of the fuel cell that realizes movement of the operating point when the determination unit determines that movement is possible (claim) 3).

Further, the battery characteristic in the poisoning state calculated by the battery characteristic calculating means and the operation region set by the operation permission region setting means are sent to the external load control means. It may be a drive device of a fuel cell provided with an information transmission means (claim 4).

Furthermore, the fuel cell drive device may be provided with safety control means for controlling the operation of the fuel cell to the safe side when the determination means determines that the operating point cannot be moved (claim). Item 5).

A catalyst poisoning rate detecting device for a fuel cell according to the present invention is a fuel cell which supplies a reaction gas to an electrode carrying a catalyst to obtain an electromotive force from a chemical reaction of the reaction gas, and an initial stage of the fuel cell. Initial battery characteristic storage means for storing in advance the battery characteristics in the state, poisoning state detection means for detecting that the catalyst is in a poisoned state, and the catalyst is poisoned state by the poisoned state detection means Output signal detection means for detecting an output signal of the fuel cell,
It is a gist to provide a poisoning rate calculation means for obtaining a poisoning rate of the catalyst from the detected output signal of the fuel cell and the cell characteristic stored by the initial cell characteristic storage means ( The thing of Claim 6).

The fuel cell system according to the present invention comprises:
A fuel cell system including a catalyst poisoning rate detection device for a fuel cell according to claim 1, wherein a load connected to the fuel cell,
Auxiliary power supply, poisoning rate of the catalyst determined by the catalyst poisoning rate detection device, poisoning rate determination means for determining whether or not a predetermined value or more, by the poisoning rate determination means When it is determined that the poisoning rate of the catalyst is greater than or equal to a predetermined value, the connection to the load is switched from the fuel cell to the auxiliary power source, and poisoning control means for executing the poisoning countermeasure for the catalyst is also provided. The provision of and is made into the gist (Claim 7
Those listed).

In the fuel cell system having the above structure, the poisoning control means includes a plurality of poisoning elimination means for eliminating poisoning of the catalyst by different methods, and the catalyst determined by the catalyst poisoning rate detection device. It is also possible to select any one of the plurality of poisoning elimination means according to the poisoning rate of, and select drive means for driving the selected poisoning elimination means. thing).

[0020]

According to the fuel cell driving apparatus of the first aspect,
When the poisoning state detecting means detects that the catalyst is in the poisoning state, the output signal detecting means detects the output signal of the fuel cell. Then, from the output signal of the fuel cell and the cell characteristics in the initial state of the fuel cell stored in advance by the initial cell characteristic storage means, the cell characteristic calculation means calculates a ratio according to the poisoning rate of the catalyst, From the ratio and the cell characteristics, the cell characteristics in the poisoned state of the fuel cell are obtained.

Therefore, according to this fuel cell driving apparatus, when the output signal of the fuel cell is detected when the catalyst is poisoned, the cell characteristics in the poisoned state can be obtained. For this reason, if the current or voltage is controlled based on the obtained battery characteristics in the poisoning state, instead of gradually groping for the current or voltage as in the conventional case, the target battery characteristics can be immediately obtained. It becomes possible to shift to.

According to the present invention, if the output of the fuel cell in the poisoned state and the cell characteristics in the initial state of the fuel cell are known, the cell characteristics in the poisoned state are uniquely determined. It was created based on new knowledge. Hereinafter, the driving method of the fuel cell will be described first, and how this new finding was revealed.

The driving method of the fuel cell is one of three control methods of constant voltage control, constant current control, and constant power control, depending on the difference between the load and the connection form and the presence / absence of a hybrid with the secondary battery. Is used. The constant power control operates at an operating point determined by the output power.
You will be driving at any operating point on the curve. However, this curve (here 0.3 W / c
It does not work on any of the m ² ) and has some limitations.

These limits are determined by the line of the minimum electrode potential for protecting the electrodes of the fuel cell, the line of the current limit determined by the gas supplyable amount of the system, and the maximum heat generation allowable amount of the system. Is the line.

Of course, since it does not exceed the curve of the initial value of the battery characteristics, it can be said that this is also a limiting condition in a certain sense (forcibly raising the gas pressure or changing the battery temperature,
Changing the initial values of battery characteristics is not discussed here). After all, this fuel cell is operated at constant power control at any operating point on the line in the curve (circle) of FIG. 1 and the range surrounded by the line (hatched region in the figure).

Since the constant voltage control and the constant current control are similar to the above description of the constant power control, detailed description thereof will be omitted here.

The cell output when the fuel cell is poisoned by a catalyst when the fuel cell is driven by the constant voltage control method will be described below. In FIG. 2, it is assumed that a fuel cell having the following cell characteristics as the initial cell characteristics is operated at the operating point A by the constant voltage control method.
Here, it is assumed that the fuel cell has its cell output lowered due to the influence of carbon monoxide (CO) poisoning of the catalyst, and its operating point has moved to B.

At this time, the information that can be detected is only one of the operating points B, but if the current density at this point is compared with the current density at the operating point A, at point B, 50 at point A is obtained.
You can see that the current of% can be passed. Therefore, at all points that compose the curve of the battery characteristics including the point A, the points at which the current becomes 50% for the same voltage are calculated, and the points obtained by these calculations are connected to draw a line. Then, the curve shown in Fig. 2 is obtained. Actually, when the cell characteristics of the fuel cell after being poisoned by CO were measured, that is, when the cell voltage was measured while changing the current, it was confirmed that the curve of FIG. did it.

That is, when the fuel cell operating under the constant voltage control has its cell output lowered due to CO poisoning,
Once the current and voltage (that is, the operating point) at a certain point after the characteristics change are determined, by comparing with the amount of current at the same voltage in the initial cell characteristics, the cell characteristics (current of the fuel cell in the CO poisoning state (current -Voltage characteristic) can be estimated.

The cell output when catalyst poisoning occurs in the fuel cell while the fuel cell is driven by the constant current control method will be described below. In FIG. 3, it is assumed that a fuel cell having the following cell characteristics as the initial cell characteristics is operated at the operating point A by the constant current control method.
Here, it is assumed that the cell output of the fuel cell is lowered due to the influence of CO poisoning and the operating point is moved to C.

At this time, the only information that can be detected is only one of the operating points C, but the voltage in the current-voltage relationship at this point is focused on, and the same voltage as this voltage is observed on the battery characteristic curve. look for. The operating point of this same voltage is point D in the figure, and comparing the current density at this point D and the current density at operating point C shows that at point C, 50% of the current at point D can flow.

Therefore, points at which the current density becomes 50% with respect to the same voltage at all points constituting the battery characteristic curve including the points A and D were calculated and obtained by these calculations. If you connect the points and draw a line, you get the curve in Figure 3. Actually, when the cell characteristics of the fuel cell after being poisoned by CO were measured, that is, when the cell voltage was measured while changing the current, it was confirmed that the curve of FIG. did it.

That is, when the fuel cell operating under constant current control has its cell output reduced due to CO poisoning,
Once the current and voltage (that is, the operating point) at a certain point after the characteristics change are determined, by comparing with the amount of current at the same voltage in the initial cell characteristics, the cell characteristics (current of the fuel cell in the CO poisoning state (current -Voltage characteristic) can be estimated.

Next, the cell output when the fuel cell is poisoned by a catalyst when the fuel cell is driven by the constant power control method will be described. In FIG. 4, it is assumed that a fuel cell having the following cell characteristics as the initial cell characteristics is operated at the operating point A by the constant power control method.
Here, it is assumed that the cell output of the fuel cell is lowered due to the influence of CO poisoning and the operating point is moved to E. Since it is a constant power control method, the movement of the operating point on the curve of the current-voltage characteristic is a curve connecting the same power output as shown in FIG.

At this time, the information that can be detected is only one of the operating points E, but the voltage in the current-voltage relationship at this point is focused on, and the same voltage as this voltage is observed on the battery characteristic curve. look for. The operating point of this same voltage is the F point in the figure, and by comparing the current density at this F point with the current density at the operating point E, it can be seen that 50% of the current at the F point can flow at the E point.

Therefore, at all points forming the curve of the battery characteristics including the points A and F, the points at which the current becomes 50% for the same voltage are calculated, and the points obtained by these calculations are obtained. If you connect and draw a line, you will get the curve in Figure 4. Actually, when the cell characteristics of the fuel cell after being poisoned by CO were measured, that is, when the cell voltage was measured while changing the current, it was confirmed that the curve of FIG. did it.

That is, when the fuel cell operating under constant power control has its cell output lowered due to CO poisoning,
Once the current and voltage (that is, the operating point) at a certain point after the characteristics change are determined, by comparing with the amount of current at the same voltage in the initial cell characteristics, the cell characteristics (current of the fuel cell in the CO poisoning state (current -Voltage characteristic) can be estimated.

From the above, when the output of the fuel cell decreases due to CO poisoning, if the current and voltage at a certain point are determined, the control method of the fuel cell is constant voltage control, constant current control, constant power control. Whichever method of CO
The cell characteristics (current-voltage characteristics) of the fuel cell in the poisoned state can be estimated.

That is, in the state of poisoning by carbon monoxide, the battery characteristics in a wide operating range can be estimated from at least one operation information. Therefore, if the current, voltage, or power control of the fuel cell is performed while the operating characteristics are foreseen after the cell characteristics are known in advance, it is possible to immediately shift to the desired cell characteristics.

As a reason why the cell characteristics (current-voltage characteristics) of the fuel cell in the CO poisoning state can be estimated, the inventor has considered the following mechanism. that is,
"A part of the platinum catalyst has stopped its function as an electrochemical reaction catalyst due to the influence of carbon monoxide. That is, the cell reaction does not occur in the portion covered with carbon monoxide, The cell reaction is occurring normally only in the non-operated area. "

CO poisoning does not change the degree of activity of all platinum catalysts of the electrodes depending on the strength of CO poisoning, but each platinum particle (at a level of only angstrom). However, if it is not a single atom), it either does not work completely or it works completely. When there is no CO poisoning, all platinum particles are working, but CO poisoning occurs Then some platinum particles stop working at all. The poisoning rate, which is the degree of CO poisoning, is determined by the ratio of the completely working platinum particles to the non-working platinum particles.

When the working condition is strong or weak, it depends on the voltage of the fuel cell, but since it is working or not working completely, the voltage of the fuel cell is the same and the current is different. Come. In other words, if all the platinum particles are working, the current that is the same as the initial value can flow, but
If 0% does not work due to CO poisoning, only 50% of the current can flow when compared at the same voltage.

New knowledge was obtained from the mechanism of CO poisoning described above. When the cell output of a fuel cell decreases due to CO poisoning, if the current and voltage (that is, operating point) at a certain point after the characteristics change are determined, compare it with the current density at the same voltage in the initial cell characteristics. The CO
It is possible to estimate the cell characteristics of a fuel cell in a poisoned state, but at the same time, from the numerical value of the percentage (%) of the current density lower than the initial value, the current point in the fuel cell It is possible to quantitatively determine the degree of CO poisoning.

As described above, the decrease in the current density can be defined by the ratio of platinum particles that do not work. Therefore, the decrease in the current density can be expressed as follows. It can be defined as "the rate at which the function stops." This ratio will be conveniently referred to as "CO coverage" with respect to the platinum catalyst. This ratio corresponds to the catalyst poisoning ratio.

Therefore, when the cell output of the fuel cell is reduced due to CO poisoning, if the current and voltage (that is, the operating point) at a certain point after the characteristic change is determined, the current density at the same voltage in the initial cell characteristics is determined. The degree of decrease in current density can be known by comparing with, and as a result, the "CO coverage" at that time in the fuel cell can be calculated. For example, in FIG. 1 to FIG. 4, if the “CO coverage” is considered, the CO coverage is 50% in all cases.

A catalyst poisoning rate detection device for a fuel cell according to a sixth aspect of the present invention calculates the "CO coverage" in accordance with the above-mentioned findings. According to the catalyst poisoning rate detecting device for a fuel cell, when the poisoning state detecting means detects that the catalyst is in the poisoning state, the output signal detecting means detects the output signal of the fuel cell. Then, from the output signal of the fuel cell and the cell characteristics in the initial state of the fuel cell stored in advance by the initial cell characteristic storage means, the poisoning rate calculation means calculates the poisoning rate of the catalyst, that is, the “CO coverage rate”. Is asked.

Therefore, according to this catalyst poisoning rate detecting device for a fuel cell, the poisoning rate of the catalyst can be obtained only by detecting the output signal of the fuel cell when the catalyst is poisoned.

The operation of the invention described in claims other than claims 1 and 6 will be described below.

According to the fuel cell driving device of the second aspect, the operation permission region setting means sets the operation region in which the operation of the fuel cell is permitted under various conditions that limit the operation of the fuel cell. Therefore, in the first aspect, the control to the cell characteristics can be made precise, and further, the operating range of the fuel cell can be known in advance from the operating region, and more precise control can be performed.

According to the fuel cell driving apparatus of the third aspect, the cell characteristic in the poisoned state is limited to the operation region set by the operation permission region setting means, and the limited cell characteristic is The determination means determines whether or not the operating point of control can be moved. When the determination means determines that the movement is possible, the control method switching permission means permits the switching of the control method of the operation of the fuel cell that realizes the movement of the operating point. Therefore, instead of gradually changing the operating point and finally moving it to a desired position as in the conventional technique,
It becomes possible to move the operating point at once by switching the operation control system, and the control speed can be improved.

According to the fourth aspect of the present invention, the fuel cell driving device is provided with the battery characteristic in the poisoning state obtained by the cell characteristic calculating means and the operation area set by the operation permission area setting means. The information transmission means sends the information to the load control means. Therefore, the load control means can control the load so as not to deviate from the operating range.

According to the fuel cell driving apparatus of the fifth aspect, the safe operation control means controls the operation of the fuel cell to the safe side when the determination means determines that the movement of the operating point is impossible. . Therefore, it is possible to quickly shift the control to the safe side rather than gradually changing the operating point to control the slow motion to the safe side.

According to the fuel cell system of the seventh aspect, when the fuel cell is likely to come out of the operation permission region, the load can be switched to the auxiliary power source before the fuel cell is actually detached, and the power source for the load can be switched. It is possible to stably supply the gas. Further, by taking measures against poisoning of the catalyst of the fuel cell, the catalyst poisoning can be eliminated.

According to the fuel cell system of the eighth aspect, any one of a plurality of poisoning elimination means is selected and selected according to the poisoning rate of the catalyst obtained by the catalyst poisoning rate detection device. The poisoning eliminating means can be driven. Therefore, in the case where the degree of poisoning of the catalyst is low, the poisoning eliminating means having relatively low responsiveness is used. In the case where the degree of poisoning is high, the poisoning eliminating means having excellent responsiveness is used. Thus, it is possible to adopt the optimum poisoning elimination means.

[0055]

Preferred embodiments of the present invention will be described below in order to further clarify the structure and operation of the present invention described above.

FIG. 5 is a schematic configuration diagram of a fuel cell power generation system 1 as a first embodiment of the present invention. As shown in FIG. 5, the fuel cell power generation system 1 uses a solid polymer fuel cell stack 10 that generates electricity, hydrogen stored in a methanol tank 12 and water stored in a water tank 14. Reformer 16 for producing rich gas
And a fuel gas supply passage 1 for sending the hydrogen-rich gas produced by the reformer 16 to the fuel cell stack 10 as a fuel gas.
8 and a fuel gas discharge passage 20 for sending the gas discharged from the fuel cell stack 10 to the outside.

Further, the fuel cell power generation system 1 is equipped with an inverter 22 for converting DC electricity generated in the fuel cell stack 10 into AC electricity.
AC electricity from 2 is not shown load (eg motor)
Supplied to The secondary battery 24 is also connected to the load, and the power supplied to the load by the switch 26 is appropriately switched between the fuel cell stack 10 on the inverter side and the secondary battery 24.

Further, the fuel cell power generation system 1 is
As an electrical control system, a carbon monoxide sensor 30 provided in the middle of the fuel gas discharge passage 20 for detecting the CO concentration in the fuel gas, the fuel cell stack 10, and the inverter 2 are provided.
2 and an output detection circuit 32 for detecting a voltage and a current amount of an electric signal output from the fuel cell stack 10 and an output signal from the carbon monoxide sensor 30 and the output detection circuit 32. The electronic control unit 80 which performs various control processes is provided.

The structure of the fuel cell stack 10 will be described below. The fuel cell stack 10 is a polymer electrolyte fuel cell as described above, and has the structure shown in FIG. 6 as its single cell structure. That is, as shown in FIG. 6, the cell has an electrolyte membrane 41, an anode 42 and a cathode 43 as gas diffusion electrodes that sandwich the electrolyte membrane 41 from both sides to form a sandwich structure, and sandwich the sandwich structure from both sides. Meanwhile, separators 44 and 45 that form fuel gas and oxidizing gas passages with the anode 42 and the cathode 43, and collector plates 46 and 47 that are disposed outside the separators 44 and 45 and serve as collector electrodes of the anode 42 and the cathode 43. It is composed of and.

The electrolyte membrane 41 is an ion exchange membrane formed of a polymer material such as a fluorine resin, and exhibits good electric conductivity in a wet state. The anode 42 and the cathode 43 are formed of a carbon cloth woven with a yarn made of carbon fiber.
Carbon powder supporting platinum as a catalyst is kneaded into the gaps of the cloth.

The separators 44 and 45 are formed of a dense carbon plate. Also, the anode 42
A plurality of ribs are formed in the separator 44 on the side, and the ribs and the surface of the anode 42 form a fuel gas flow channel groove 44p. On the other hand, a plurality of ribs are also formed on the separator 43 on the cathode 43 side, and the ribs and the surface of the cathode 43 form a flow channel 45p for the oxidizing gas. The current collectors 46 and 47 are formed of copper (Cu).

Although the structure of the single cell of the fuel cell stack 10 has been described above, the separator 44,
A plurality of sets of the anode 42, the electrolyte membrane 41, the cathode 43, and the separator 45 are laminated in this order, and the current collector plate 4 is provided outside thereof.
By disposing 6, 47, the fuel cell stack 10
Is composed.

The fuel gas supply passage 18 connects the reformer 16 and the anode side gas inlet 10a (FIG. 5) of the fuel cell stack 10. In practice, the anode side gas inlet 10a is connected to a manifold (not shown). The fuel cell stack 10 is branched and connected to the plurality of flow channel grooves 44p on the fuel gas side through the manifold. On the other hand, the anode side gas outlet 10b of the fuel cell stack 10 is connected to a manifold (not shown),
Through this manifold, the fuel cell stack 10 is branched and connected to a plurality of flow channel grooves 44p (connected from the side opposite to the fuel gas supply passage 18).

The structure of the carbon monoxide sensor 30 will be described below. FIG. 7 is a vertical cross-sectional view of the carbon monoxide sensor 30. As shown in FIG. 7, this carbon monoxide sensor 3
0 is an electrolyte membrane 50 and two electrodes 52 and 54 that sandwich the electrolyte membrane 50 from both sides to form a sandwich structure,
By sandwiching this sandwich structure from both sides, two mesh-shaped metal plates 5 for preventing the sandwich structure from bending
6, 58 and this sandwich structure and metal plate 56,
Two holders 60 and 62 for holding 58 and both holders 6
Insulating member 64 for connecting 0 and 62 in an electrically insulated state
With.

The electrolyte membrane 50 is a proton conductive membrane formed of a solid polymer material such as a fluorine resin. The electrodes 52, 54 are made of carbon cloth woven from threads made of carbon fibers, and carbon powder carrying platinum as a catalyst is kneaded into the spaces between the cloth.

Specifically, the electrolyte membrane 50 and the electrodes 52 and 54 are joined by the following method.

On the surface of the electrode base material (carbon cloth or carbon paper), a catalyst powder prepared by previously supporting platinum on the surface of carbon powder is applied, and the electrolyte membrane 50 and this electrode base material are integrated by hot pressing. What to do. On the surface of the electrode base material, a catalyst powder prepared by previously supporting platinum on the surface of carbon powder is applied, and the electrolyte membrane 50 and this electrode base material are bonded and integrated with a proton conductive solid polymer solution. thing.

A catalyst powder prepared by previously supporting platinum on the surface of carbon powder is dispersed in an appropriate organic solvent to form a paste, which is applied to the surface of the electrolyte membrane 50 by a method such as screen printing. After that, it is integrated with the electrode substrate by hot pressing. On the surface of the electrolyte membrane 50, sputtering method, vapor deposition method, CVD
Thin film forming method such as the PVD method or PVD method, which carries platinum and then is integrated with the electrode base material by hot pressing.

The carbon powder supporting the platinum catalyst is prepared by the following method. An aqueous solution of platinum sulfite complex is obtained by mixing an aqueous solution of chloroplatinic acid and sodium thiosulfate. While stirring this aqueous solution, the hydrogen peroxide solution is removed to deposit colloidal platinum particles in the aqueous solution. Next, carbon black as a carrier [eg VulcanX
While adding C-72 (trademark of CABOT Inc. in the United States) and Denka Black (trademark of Denki Kagaku Kogyo Co., Ltd.),
With stirring, colloidal platinum particles are attached to the surface of the carbon black. Next, after suction filtration or pressure filtration of the solution to separate the carbon black to which the platinum particles are attached,
After being repeatedly washed with deionized water, it is completely dried at room temperature. Next, the agglomerated carbon black is pulverized by a pulverizer, and then, in a hydrogen reducing atmosphere, at 250 ° C. to 350 ° C. for 2 hours.
By heating for about an hour, the platinum on the carbon black is reduced and residual chlorine is completely removed, and the platinum catalyst is completed.

The platinum catalyst used for the electrolyte membrane 41 of the fuel cell stack 10 described above is also formed by the same method. The area of the electrodes 52 and 54 is
0.1cm ² ~1cm ² about is desirable.

The metal plates 56 and 58 are mesh-shaped and have a structure that does not prevent the gas from reaching the electrodes 52 and 54. As a material thereof, a material having excellent electric conductivity, hardly rusting, and not causing hydrogen embrittlement is desirable, and specifically, titanium or stainless steel is used. Alternatively, the surface of the mesh-shaped copper plate may be coated (eg, plated) with a metal such as gold, platinum, or titanium. Further, as long as it satisfies the above-mentioned required performance, the surface of a porous carbon plate or foamed nickel is coated (for example, plated) with a metal such as gold, platinum, or titanium, or the surface of an engineering plastic is coated with gold. It may be coated (for example, plated) with a metal such as platinum, titanium, or the like to ensure electrical conductivity.

The holders 60, 62 have a shape having flanges 60a, 62a inside the cylinder.
The electrolyte membrane 50, the electrodes 52 and 54, and the metal plate 5 at 62a.
Hold 6, 58. As the material, a material that has excellent electric conductivity, is hard to rust, and does not cause hydrogen embrittlement,
Specifically, titanium or stainless steel is used. Alternatively, a copper plate whose surface is coated (for example, plated) with a metal such as gold, platinum, or titanium may be used. Furthermore, as long as it satisfies the above-mentioned required performance, a dense carbon plate or a surface of an engineering plastic is coated (for example, plated) with a metal such as gold, platinum, or titanium to secure electrical conductivity. It doesn't matter.

Incidentally, on the electrolyte membrane 50 side of the holder 62,
An O-ring 66 is provided to prevent the atmosphere on one electrode side from leaking to the other electrode side. Where O
Instead of the ring 66, the end of the electrolyte membrane 50 is replaced by the holder 62.
It is also possible to adopt a configuration in which the sealing property is ensured by directly applying an adhesive or by thermocompression.

Screws 60 are provided on the outer circumference of the holders 60 and 62.
b and 62b are cut, and these screws 60b and 62b are
And two screws 64a cut inside the insulating member 64,
Both holders 60, 62 are screwed together with 64b.
Are connected while sandwiching the electrode 52, the electrolyte membrane 50 and the electrode 54 therebetween. As a material of the insulating member 64, for example, Teflon is used.

Further, the carbon monoxide sensor 30 has a gas inflow passage 68 connected to the holder 60 on one side by screwing. The gas inflow passage 68 is a passage for guiding the gas to be detected to the electrode 52, and is made of an insulating material. No special gas passage is connected to the holder 62 on the other side, and the electrode 54 is open to the atmosphere.

Further, this carbon monoxide sensor 30 has detection terminals 60T and 62T provided on both holders 60 and 62.
And an electric circuit 70 for electrically measuring the electromotive force generated between the electrodes 52 and 54. This electric circuit 7
0 is composed of a voltmeter 72 and a load current adjusting resistor 74, and the signal of the voltmeter 72 is output to an external control system. The voltmeter 72 is connected so that the detection terminal 60T of the holder 60 on the electrode 52 side to which the fuel gas is supplied has a negative pole, and the detection terminal 62T of the holder 62 on the electrode 54 side communicating with the atmosphere has a positive pole. There is.

The carbon monoxide sensor 30 thus constructed
Is connected to the branch port 18a of the fuel gas supply passage 18 by screwing, and is used for detecting the CO concentration in the fuel gas supplied to the fuel cell main body (not shown).

In this carbon monoxide sensor 30, the electrode 52
When the fuel gas is supplied to the electrode 5 through the electrolyte membrane 50
Since an electromotive force is generated between 2 and 54, this electromotive force is detected by using the voltmeter 72 of the electric circuit 70. Since the electromotive force is reduced due to the poisoning of the catalyst by carbon monoxide as in the fuel cell main body, it is small when the CO concentration is high and is large when the CO concentration is low. Therefore, it is possible to measure the CO concentration of the gas to be detected by previously examining the relationship between the CO concentration and the measured value of the voltmeter 72 at that time using a gas having a known CO concentration. Since the detection sensitivity during this measurement is not affected by hydrogen, it is necessary to measure carbon monoxide in the detected gas containing a large amount of hydrogen such as fuel gas with high accuracy. You can

The electronic control unit 80 is constructed as a logic circuit centering on a microcomputer, and more specifically, a CPU 82 for executing a predetermined calculation and the like according to a preset control program, and various calculation processes by the CPU 82. A ROM 84 in which control programs and control data necessary for the above are stored in advance, a RAM 86 in which various data necessary for executing various arithmetic processes in the CPU 82 are temporarily read and written, a carbon monoxide sensor 30, and output detection The output signal from the circuit 32 and the output signal from the electronic control unit (not shown) for the load are input, and the reformer 16, the fuel cell stack 10, the inverter 22 and the switching device are input according to the calculation result of the CPU 82. 26 is provided with an input / output interface 88 for outputting a control signal.

The ROM 84 stores in advance the initial cell characteristic data DT0 indicating the cell characteristics of the fuel cell stack 10 in the initial state. FIG. 8 shows an example of the initial battery characteristic data DT0. As shown in FIG. 8, the initial battery characteristic data DT0 is a two-dimensional map showing the characteristic L0 determined by the current density and the voltage, and the voltage decreases as the current density increases. Current density is 200
When [mA / cm ^2], the voltage is 0.80 [V], and the voltage at a current density of 400 [mA / cm ^2] and has a 0.73 [V].

Cell characteristics L in the initial state of the fuel cell
0 is controlled so that the operating conditions such as the gas pressure, the gas utilization rate, and the cell temperature are always constant if the fuel cell is in steady operation. Therefore, the cell characteristic L0 in such a state is unique. Can be specified. Therefore, the initial battery characteristic data DT0 is stored as one data.

Further, in FIG. 5, only the gas system on the anode side is shown for the gas system, and the gas system on the cathode side is omitted.

C of the electronic control unit 80 having such a configuration
By the PU 82, the reformer 16 and the fuel cell stack 1
0, the inverter and the switch 26 are controlled appropriately. As a result, in the fuel cell power generation system 1, proper electric power is supplied to the load while selecting the optimum one from the three types of control methods of constant voltage control, constant current control, and constant power control.

Considering first when the load is not operating (for example, when it is stopped in the case of an electric vehicle), the fuel cell stack 10 is under constant voltage control at this time, and the fuel cell stack 10 is operated at a constant voltage. Electricity is supplied to the secondary battery 24. After that, when the load starts to move (when traveling in an electric vehicle), the fuel cell stack 10 is operated under constant power control with the electric power specified by the load,
The switch 26 is switched to supply the electricity from the fuel cell stack 10 to the load. Here, in the case where the control system is switched from the constant voltage control to the constant power control to supply electricity, C of the electronic control unit 80 is
How the control process is executed by the PU 82 will be described.

9 to 10 show an electronic control unit 8
It is a flow chart which shows a control routine performed by 0 of CPU82. This control routine is executed after receiving an instruction from the electronic load control unit that the load has started to operate, and is repeatedly executed at predetermined time intervals. As shown in FIG. 9, when the processing is started, the CPU 82 first determines from the output signal from the electronic control unit for load whether or not the load is constant (step S100), and further, the fuel cell. It is determined whether or not the gas supply state (gas amount, etc.) to the stack 10 is constant (step S110).

The CPU 82 executes step S100 or S
If a negative determination is made at 110, the process proceeds to "return" to end this control routine once. On the other hand, if the positive determination is made in both step S100 and step S110, that is, if the load is constant and the gas supply state is constant, the process proceeds to step S120.

In step S120, the CPU 82 sets the output voltage from the voltmeter 72 of the carbon monoxide sensor 30 to C
The process of inputting as the O concentration D is performed. Next, it is determined whether or not the carbon monoxide concentration D is higher than a predetermined concentration D0 set in advance, thereby performing a process of determining whether or not the catalyst of the anode electrode is poisoned (step S130). ).

When it is determined in step S130 that the catalyst is not in the poisoning state, the CPU 82 "returns" the processing.
To end the control routine. On the other hand, if it is determined in step S130 that the catalyst is poisoned, the current voltage V1 and current I1 of the fuel cell stack 10 detected by the output detection circuit 32 are input (step S140).

Then, its current voltage V1 and current I
1 and the initial battery characteristic data DT0 shown in FIG. 8 stored in the RAM 86 in advance, the current battery characteristic data DT
Is calculated (step S150). This processing is specifically performed as follows. At this point in time, since the fuel cell stack 10 is operating under constant voltage control, first, the characteristic L determined by the initial cell characteristic data DT0 is set.
On 0, the current value I0 corresponding to the current voltage V1 is obtained. Specifically, the map of the initial battery characteristic data DT0 stored in the ROM 84 is directly copied to the RAM 86, and the current value I0 is obtained on the map of the RAM 86 (see FIG. 11).

Next, the ratio R between the current value I0 and the current I1 (which is a ratio corresponding to the poisoning rate of the catalyst) is obtained, and each point on the initial battery characteristic L0 is multiplied by the ratio R. By doing so, the current battery characteristic L1 is obtained. This calculation is also performed on the map stored in the RAM 86,
The current value at each point on the initial battery characteristic L0 is multiplied by the ratio R to obtain a line reduced in the axial direction of the current density by the ratio, that is, the current battery characteristic L1 (see FIG. 11). ).

Subsequently, the CPU 82 proceeds to step S150.
On the map of the current battery characteristic data DT1 calculated in
A process of setting an operation permission area in which the operation of the controlled operation is permitted is performed (step S160). The hatched area shown in FIG. 1 corresponds to this operation permission area. In particular,
As shown in FIG. 12, a line G1 of, for example, 500 mV determined by limiting the minimum electrode potential for protecting the electrodes of the fuel cell, and a line G2 of, for example, 700 mA / cm ² of current limitation determined by the gas supply capacity of the system. , Determined by the maximum heat generation capacity of the system, for example, 0.6 W / cm ²
Line G3 (here: fever is originally "J: Joule")
However, for the sake of clarity, the area surrounded by W) and the battery characteristic L0 in the initial state (hatched area in the figure) is set as the operation permission area E.

After execution of step S160, the CPU 82
Advances the processing to step S170 of FIG.
A control signal is sent to the control unit 6 to switch the load connection to the fuel cell stack 10 side (step S17).
0).

Next, the CPU 82 proceeds to step S160.
A process of determining whether or not the constant power control system can be operated in the operation permission region E calculated in step S180 is performed (step S180). This constant power control operation is controlled by constant power defined by the degree of load, and whether or not this can be executed in the operation permission area E is determined as follows. As shown in FIG. 13, first, on the map (shown in FIG. 12) obtained after execution of step S160, a line L2 of constant power determined according to the degree of load is written, and this line L2 and step S15 are written.
An intersection P with the current battery characteristic L1 obtained by 0 is obtained. This intersection point P is the operating point to be operated by constant power control,
It is determined whether or not this intersection P is located within the operation permission area E set in step S160. In this way, it is determined whether or not it is possible to shift from the current constant voltage control method operation to the constant power control method operation.

When it is determined in step 180 that the intersection P exists within the operation permission area E, that is, it is possible to shift to the operation of the constant power control system, the CPU 82
As shown in FIG. 14, the current density is increased at once from the current operating point Q to a new operating point (intersection P), and constant power operation is executed (step S190). Then the CPU
Reference numeral 82 advances the process to "return" to temporarily end this control routine.

On the other hand, if it is determined in step S180 that the intersection P does not exist in the operation permission area E, that is, it is impossible to shift to the operation of the constant power control system, the fuel cell stack 10 is safely operated. Processing for switching to the operation is executed (step S195). This safe operation is performed by stopping the operation of the fuel cell stack 10, controlling such that the CO concentration in the fuel cell gas from the reformer 16 is reduced, or continuously operating at constant voltage without switching to constant power operation. Or a control for reducing the amount of electric power on the load side is employed. After the processing of step S195, the CPU 82 advances the processing to “return” and once ends this control routine.

According to the fuel cell power generation system 1 of the first embodiment configured as described above, the catalyst is poisoned by CO under the constant voltage control, and the operating point changes from the O point (in the initial state). Q
When the constant power control operation is performed after the change to the point, the operating point can be immediately moved to the point P to obtain a predetermined power.

In the conventional technique, as shown in FIG. 15, when the constant power control is operated after the operating point is changed from the O point to the Q point, the current is gradually changed by Δi and finally the P point is reached. It reached to and was getting a predetermined power. Therefore, it takes a lot of time to shift the operating point to the point P, and as a result,
Control delay occurred in battery output. On the other hand, according to the fuel cell power generation system 1 of this embodiment, as described above, the operating point can be immediately moved to the point P to obtain the predetermined electric power, so that the control delay as described above may occur. Nothing.

In the prior art, the amount of change in current Δi
If is too large, the control may overshoot, and as a result, the fuel cell stack 10 may stall itself. On the other hand, according to this fuel cell power generation system 1, since it is possible to shift to a desired operating point with high accuracy, such a stall does not occur.

As another example, a case where the operating point is changed from the O point to the K point due to CO poisoning as shown in FIG. 16 in the prior art will be described. The poisoning rate of the catalyst at this time is assumed to be much higher than that shown in FIG. After this, when the operation shifts to constant power control, the current is
However, since it falls below the limit G1 of the minimum electrode potential of the fuel cell before the predetermined output power is obtained, a safe operation works at the operating point M, and the fuel cell stops operating. I will end up.

On the other hand, in the fuel cell power generation system 1 of this embodiment, as shown in FIG. 17, after the operating point changed from the O point to the K point due to CO poisoning, information on the initial cell characteristics L0 was obtained. And the current output, that is, the value of the point K, the current battery characteristic L1 can be obtained, and at this time, the operating point for obtaining the predetermined power is no longer within the limits of G1 to G3 at this point. It is possible to know that the constant power control cannot be operated. Therefore,
Based on such information, the safe operation can be immediately actuated, and there is no delay in controlling the safe operation.

In this embodiment, the operation of the fuel cell stack 10 is predicted before the operation is out of the operation permission region to operate the safe operation. At this time, the switch 26 is switched to change the load. Is preferably connected to the secondary battery 24. In this case, from the viewpoint of the load, the electric power is supplied stably without interruption. Therefore, the controllability of the entire system can be improved.

In the above-mentioned embodiment, the control in the case of switching the operation system of the fuel cell from the constant voltage control to the constant power control has been described in detail, but the same applies to the case of switching from the constant voltage control to the constant current control. Can be adapted to. That is, similar to the above-described embodiment, the current value (operating point) of the constant current control to be switched is calculated after obtaining the current battery characteristic L1 in the state where the catalyst is poisoned.
It is checked in advance whether or not the current battery characteristics within the operation permission region can be taken, and if possible, the operation point is changed at once. With such a configuration, even when the constant voltage control is switched to the constant current control, proper operation control can be performed without control delay or stall of the fuel cell.

Further, a case will be described in which the operation system before switching the fuel cell is not the constant voltage control as described above, but another control, that is, constant current control or constant power control. When the fuel cell stack 10 is driven by the constant current control method, the cell characteristics when the catalyst poisoning occurs in the fuel cell stack 10 are as described with reference to FIG. The current voltage V1 and current I1 after poisoning (which decreases at a constant current) are detected, and the same voltage as this voltage V1 is searched for on the battery characteristic L0 in the initial state. The ratio R between the current density and the current I1 at the searched point is obtained, and each point on the battery characteristic L0 in the initial state is multiplied by the ratio R to obtain the current battery characteristic L1.

In addition, the cell characteristics when the catalyst poisoning occurs in the fuel cell stack 10 while the fuel cell stack 10 is driven by the constant power control method are described in the section of action in FIG.
As described with reference to FIG.
1. A current I1 (which decreases at a constant power) is detected, and the same voltage as the voltage V1 is searched for on the battery characteristic L0 in the initial state. The ratio R between the current density and the current I1 at this searched point is obtained, and each point on the battery characteristic L0 in the initial state is calculated as the ratio R.
The current battery characteristic L1 is obtained by multiplying by.

After obtaining the current battery characteristic L1 in this way, as described above, it is determined whether or not the operating point in the control method to be switched can be set on the current battery characteristic in the operation permission region. Make sure to check in advance. With this configuration, when switching the control method,
Appropriate operation control can be performed without control delay or stall of the fuel cell.

In the above embodiment, the carbon monoxide sensor 30 is used to detect the CO in the fuel gas as the poisoning state detecting means.
When the concentration was detected and the CO concentration became equal to or higher than a predetermined concentration, the catalyst was determined to be in a poisoned state. However, the present invention is not limited to this.
Another configuration, for example, a configuration using a potentiostatic carbon monoxide sensor may be used. Alternatively, the temperature difference between the fuel gas inflow side and the fuel gas outflow side of the electrode may be detected, and the poisoning state of the catalyst may be estimated from the temperature difference to detect the poisoning state of the catalyst.

The second embodiment of the present invention will be described below. In the second embodiment, as shown in the schematic configuration diagram of FIG. 18, the fuel cell power generation system 1 described in the first embodiment is described.
To the entire fuel cell system including the load system 201 that receives the supply of electricity from the fuel cell power generation system 1.

The load system 201 is a load (here, a motor) 2 connected to the output of the fuel cell power generation system 1.
10 and a load electronic control unit 220 that controls the operation of the load 210.

The load electronic control unit 220 is configured as a logic circuit centering on a microcomputer, like the electronic control unit 80 on the fuel cell power generation system 1 side (hereinafter referred to as the fuel cell electronic control unit) 80. CP
A U222, a ROM 224, a RAM 226 and an input / output interface 228 are provided. The load 210 is electrically connected to the input / output interface 228,
By sending a control signal to the load 210, the operation of the load 210 can be controlled. Further, the input / output interface 228 is electrically connected to the input / output interface 88 of the fuel cell electronic control unit 80, and data is exchanged between the load electronic control unit 220 and the fuel cell electronic control unit 80. Communication is possible.

What kind of data is exchanged between the electronic control units 80 and 220 and the control process is executed will be described below. FIG. 19 is a flowchart showing both control routines executed by the CPUs 82 and 222 of the electronic control units 80 and 220.

As shown in FIG. 19, when the CPU 82 of the fuel cell electronic control unit 80 starts processing,
First, the same processing as steps S100 to S160 (FIG. 9) of the first embodiment is executed. Then, step S15
The current battery characteristic data DT1 calculated at 0 and the data DT2 indicating the operation permission area E calculated at step S160
Are output from the input / output interface 88 to the load electronic control unit 220 (step S300).

On the other hand, the load electronic control unit 22
In the CPU 222 of 0, the fuel cell electronic control unit 8
The current battery characteristic data DT1 and the data DT2 indicating the operation permission region E sent from 0 are input (step S40
0), it is determined whether or not the power consumption defined by the load 210 is appropriate as viewed from both data DT1 and DT2 (step S).
410). In detail, the process of step S410 generates the map shown in FIG. 12 from the current battery characteristic data DT1 and the data DT indicating the operation permission area E, and the following process is performed on this map.

That is, as shown in FIG. 20, the load 210 obtained according to the degree of the load 210 on the same map as in FIG.
The power consumption line L2 is written, the intersection point P between this line L2 and the line L1 indicated by the current battery characteristic data DT1 is obtained, and it is checked whether or not this intersection point P exists in the operation permission area E. Here, if it is determined that the intersection point P exists within the operation permission area E, it is determined that the power consumption defined by the load 210 is appropriate, while the intersection point P exists outside the operation permission area E. When it is known (in the case of the intersection P1 in the figure), it is determined that the power consumption is inappropriate.

When it is determined in step S410 that the power consumption of the load 210 is inappropriate, the CPU 222 sets the power consumption of the load 210 to the current size (for example, L in the figure).
2'line) to the size indicated by line L3, and the line L3 and the current battery characteristic data DT1
The intersection P2 with the line L1 indicated by is included in the operation permission area E (step S420). The intersection P3 at this time is a point where the current becomes the largest in the operation permission region E.

After execution of step S420, or when it is determined in step S410 that the power consumption of the load 210 is appropriate, the process proceeds to step S430, and the load 210 is sent to the fuel cell electronic control unit 80. Input / output interface 2 sends a signal to the effect that switching is OK
28 to the fuel cell electronic control unit 80.

On the side of the fuel cell electronic control unit 80,
Wait for the signal "OK" to be sent from the load electronic control unit 220 side (step S310),
The process proceeds to subsequent step S320. Step S32
At 0, a control signal is sent to the switch 26 to switch the connection of the load 210 to the load electronic control unit 220 side (step S310), and a new operating point Q after catalyst poisoning is newly added. The current density is increased at once to the operating point P, and constant power operation is executed (step S22).
0). Note that the operating point P referred to here is the step S420.
When the processing for reducing the power consumption of the load is executed at, the position is the intersection P2 in FIG. After that, CPU8
2 advances the processing to "return" and once ends this control routine.

In the second embodiment described in detail above, the current cell characteristic data DT1 calculated by the electronic control unit 80 on the fuel cell power generation system side and the data DT2 indicating the operation permission region E are stored on the load side electronic control unit. It is configured to send to 220. Therefore, the load-side electronic control unit 220 can control the magnitude of the load so as not to deviate from the operation permission region E. Therefore, on the side of the fuel cell power generation system, it is not necessary to individually judge whether or not the operating point to be switched is included in the operation permission region E, and the operation control system can be switched at a stroke, which is excellent in controllability. There is an effect of being.

A third embodiment of the present invention will be described.
Similar to the second embodiment, the third embodiment relates to the entire fuel cell cell system including the fuel cell power generation system 1 and the load system 201, and the hardware configuration is almost the same as that of the second embodiment. It is a thing. The difference is that the ROM 84 of the fuel cell electronic control unit 80
The poisoning elimination method menu MN described later is stored in advance. The control routine executed by the fuel cell electronic control unit 80, which is software, is also different from that of the second embodiment. Hereinafter, this control routine will be described in detail with reference to the flowcharts of FIGS. The control routine is repeatedly executed at predetermined time intervals.

As shown in FIG. 21, when the CPU 82 of the fuel cell electronic control unit 80 starts processing,
First, the same processing as steps S100 to S140 (FIG. 9) of the first embodiment is executed. Then, step S14
The voltage V1 and the current I1 fetched at 0 and the RAM86
A process of calculating the poisoning rate R of the catalyst is performed from the initial battery characteristic data DT0 stored in advance (step S50).
0). Specifically, since the fuel cell is operated by the constant voltage control at this time point, first, the current voltage V on the characteristic L0 determined by the initial cell characteristic data DT0.
A current value I0 corresponding to 1 is obtained. Next, the ratio R between the current value I0 and the current I1 is calculated and used as the poisoning rate of the catalyst. The specific contents of step S500 will be described in step S15 of the first embodiment described above.
Since the processing is 0, detailed description is omitted here.

After execution of step S500, the CPU 82
Advances the process to step S510 shown in FIG. In step S510, a process of determining whether or not the poisoning rate R of the catalyst obtained in step S500 is equal to or greater than a predetermined value R0 is performed.

The poisoning rate R of this catalyst indicates the CO coverage which is the proportion of platinum particles forming the catalyst whose surface is coated with carbon monoxide, and the poisoning rate R is large. Therefore, the poisoning state can be quantitatively known. Therefore, in step S510, it is possible to determine with high accuracy whether or not the catalyst is in the poisoning state by determining whether or not the poisoning rate R is equal to or greater than the predetermined value R0.

At step 510, the poisoning rate R is the predetermined value R0.
If the above is the case, that is, if it is determined that the catalyst is in a poisoned state, the CPU 82 advances the process to step S520. In step S520, CPU 82 sends a control signal to switch 26 to perform a process of switching the connection of load 210 to the side of secondary battery 24. Then, the poisoning elimination method menu MN stored in the ROM 84 is searched based on the poisoning rate R of the catalyst obtained in step S500, and a predetermined poisoning countermeasure determined by the poisoning rate R is selected (step). S530).

An example of the poisoning elimination method menu MN is shown in FIG. As shown in FIG. 23, the poisoning elimination method menu MN is a table including an A column of the poisoning rate R of the catalyst, which is a search key item, and a B column indicating a poisoning elimination method.
In step S520, the poisoning rate R calculated in step S500 is compared with the column A to specify the column B and select its content. In the column B, a first method of increasing the temperature of the reforming reaction section of the reformer 16, a second method of decreasing the temperature of the oxidation reaction section of the reformer 16, and the reformer The third method of increasing the amount of air supplied to the 16 oxidation reaction parts is stored. Note that these methods have higher responsiveness as the rank becomes lower, and the corresponding poisoning rate R becomes large.

Contents of column A of the poisoning elimination method menu MN and B
The contents of the column are determined based on the following ideas. When the poisoning rate R is small, there is time to allow the fuel cell to stall due to carbon monoxide poisoning, so the time from the start of control to the effect may be long (rather, It is desirable to select and implement a control method that has less side effects such as lowering the energy efficiency of the system due to the control). On the other hand, when the poisoning rate R is large, there is no time to allow the fuel cell to stall due to carbon monoxide poisoning, and therefore it is necessary to perform control that immediately exerts an effect (the control causes energy consumption of the system). Even if there are side effects such as reduced efficiency, it is desirable to select and implement a control method that is effective immediately. However, since the menu of the poisoning elimination method is determined by the system configuration of each fuel cell, it is necessary to determine the poisoning elimination menu in advance according to the system configuration of each fuel cell.

After execution of step S530, the CPU 82
Executes the poisoning countermeasure selected in step S530 (step S540). When these measures are taken, the carbon monoxide concentration in the reformed gas supplied from the reformer 16 to the fuel cell stack 10 is reduced, and as a result, the poisoning of the catalyst is recovered. After that, the CPU 82 sets a value 1 to the flag F indicating the connection state of the load 210, and then “return”.
Then, the control routine is ended once.

On the other hand, in step S510, if it is determined that the poisoning rate R of the catalyst is smaller than the predetermined value R0, that is, the catalyst is not in the poisoning state, the process proceeds to "return" in principle. This control routine is once ended. In fact, when the negative determination is made in step S510, it is determined whether or not the above-mentioned flag F has the value 1 (step S56).
0), if it is determined that the flag F has a value of 1, the connection state of the load switched to the secondary battery 24 side in step S520 is returned to the fuel cell stack 10 side (step S57).
0), the flag F is cleared to the value 0 (step S58).
0). That is, as a result of taking countermeasures against poisoning in step S540,
When the poisoning rate R of the catalyst becomes smaller than the predetermined value R0, it is determined that the CO poisoning of the fuel cell stack 10 has been recovered, and the load 210 is switched to the fuel cell stack 10 side.

In the third embodiment described in detail above, the poisoning state of the catalyst is quantitatively obtained as the poisoning rate R by detecting the output signal of the fuel cell stack 10 when the catalyst is poisoning state. You can Further, when the determined poisoning rate R becomes a predetermined value R0 or more, the connection to the load is switched from the fuel cell stack 10 to the secondary cell 24, and the catalyst poisoning countermeasure is executed to stabilize the load. It is possible to effectively supply power and eliminate the poisoning state of the catalyst.

Further, in this embodiment, one of a plurality of poisoning elimination methods is selected according to the obtained catalyst poisoning rate R, and therefore, depending on the degree of catalyst poisoning. You can select the appropriate one. When the poisoning rate R of the catalyst is low, it is possible to select the poisoning eliminating means having relatively poor responsiveness, and when the poisoning rate R is high, the poisoning eliminating means having excellent responsiveness may be selected. it can. Since the poisoning elimination means having poor responsiveness generally causes little damage to the fuel cell stack 10, the poisoning elimination means is selected according to the poisoning rate R to reduce the damage to the fuel cell stack 10. At the same time, the poisoning state can be eliminated.

In the third embodiment, the poisoning rate R of the catalyst is quantitatively calculated, and when the poisoning rate R exceeds a predetermined value, the load is connected to the secondary battery 24. Although the poisoning of the catalyst was eliminated while switching to the side, the calculated poisoning rate R of the catalyst may be used as follows instead. Characteristics having a proportional relationship with the poisoning rate R (for example, the current-voltage characteristics and the current-power characteristics described above) have been measured in advance by experiments (for example, the poisoning rate of 10%, 20
%, 30%, 40%, 50%, and the like, the result that these characteristics are obtained is measured in advance.
Store in M. Then, in accordance with the poisoning rate R obtained by the above calculation, an appropriate one of the above characteristics is selected from the stored contents of the ROM. With such a configuration, if the poisoning rate R is known, the possible characteristics can be predicted, so that the control response can be improved.

As an example of the control characteristic proportional to the poisoning rate R, the heat generation from the polymer electrolyte fuel cell will be taken as an example and described in detail below. As shown in FIG. 24, the calorific value from the polymer electrolyte fuel cell increases as the poisoning rate R increases with the same current density. Conventionally, while controlling the fuel cell, the temperature of the main body of the fuel cell, the cooling water temperature of the fuel cell, the gas (anode and cathode) outlet temperature of the fuel cell, etc. are measured, and based on these measurement results, the fuel The temperature of the fuel cell is kept constant by adjusting the cooling amount of the cell (circulation amount of cooling water).

The fundamental idea of such a control device is that the control is performed only after the fact that the temperature has changed is known, and the control is not performed by anticipating the change in temperature. Therefore, it takes time for the temperature of the fuel cell to actually change, to detect it, to perform control to return the temperature, and to actually return the temperature of the fuel cell to the original value. During that time, the fuel cell is out of the prescribed temperature,
So to speak, it is an unstable operation. Of course, there are various control methods to shorten the control delay as much as possible, but the more the control method is devised, the more complicated the control content (both hardware and software) becomes.
As a result, the manufacturing cost of the fuel cell system is increased.

On the other hand, according to the example of the present invention, if the poisoning rate R is known, the heat generation amount at each operating point can be known. Therefore, when the operating point is changed at a certain poisoning rate R, the calorific value of the fuel cell at the new operating point can be calculated in advance. For example, when increasing the current density, the amount of heat generated from the fuel cell can be known in advance, so the circulation amount of the cooling water can be increased. Therefore, there is an effect that the control delay described above can be reduced to a level where there is practically no problem.

In the above-described first to third embodiments, the case where platinum is used as the catalyst of the anode 42 of the fuel cell stack 10 has been described, but in addition to this, it is the first component as the anode side electrode catalyst. An alloy catalyst composed of an alloy of platinum and one or more components of ruthenium, nickel, cobalt, vanadium, palladium, indium, etc. as the second component may be used. Also in this case, the same effects as those of the first to third embodiments can be obtained.

The embodiments of the present invention have been described above.
The present invention is not limited to these examples, and it goes without saying that the present invention can be implemented in various modes without departing from the scope of the present invention.

[0135]

As described above in detail, in the fuel cell drive apparatus according to the first aspect of the present invention, when the output of the fuel cell is lowered due to carbon monoxide poisoning of the catalyst, a certain current and voltage are applied. If determined, the cell characteristics (current-voltage characteristics) of the fuel cell in the poisoned state can be estimated. Therefore, if the battery characteristics are known in advance and the operation control is performed while predicting the operation characteristics, it is possible to immediately shift to the desired battery characteristics. Therefore, proper operation control can be performed without a control delay or a stall of the fuel cell.

In particular, in the fuel cell drive apparatus according to the second aspect, the operating range of the fuel cell can be known in advance from the operating region in which the operation of the fuel cell is permitted, so the control accuracy can be improved.

According to the fuel cell driving apparatus of the third aspect,
Instead of gradually changing the operating point to finally move it to the desired position as in the conventional technique, the operating point can be changed at a stroke by switching the operation control method, which improves the control speed. Can be planned.

According to the fourth aspect of the fuel cell drive system of the present invention,
By sending the battery characteristics in the poisoning state and the operating area set by the operation permission area setting means to the external load controlling means, the load controlling means does not deviate from the operating area. Control can be performed. For this reason, in this fuel cell drive device, it is not necessary to make a determination for each operating point that deviates from the operating region, and controllability is excellent.

According to the fuel cell driving apparatus of the fifth aspect,
When the movement of the operating point is impossible, the operation of the fuel cell is controlled on the safe side. Therefore, rather than gradually changing the operating point to control the loose motion to the safe side, the control can be quickly shifted to the safe side, which is excellent in safety.

In the fuel cell catalyst poisoning rate detecting device according to the sixth aspect, the catalyst poisoning rate can be quantitatively obtained by detecting only the output signal of the fuel cell when the catalyst is poisoned.

In the fuel cell system according to the seventh aspect, when the fuel cell is likely to come out of the operation permission area,
The load can be switched to the auxiliary power source before it is actually removed, and the power can be stably supplied to the load. Further, by taking measures against poisoning of the catalyst of the fuel cell, it is possible to eliminate the catalyst poisoning.

In the fuel cell system according to the eighth aspect, when the poisoning rate of the catalyst is low, the poisoning eliminating means has relatively poor responsiveness, and when the poisoning rate is high, the response is reduced. It is possible to employ the optimum poisoning elimination means such as the poisoning elimination means having excellent properties. Poisoning elimination means with poor responsiveness generally causes less damage to the fuel cell,
By selecting the poisoning elimination means according to the poisoning rate, the poisoning state can be eliminated while reducing the damage to the fuel cell.

[Brief description of drawings]

FIG. 1 is a graph showing limitations on battery characteristics.

FIG. 2 is a graph showing how, when catalyst poisoning occurs in a fuel cell under the operation of constant voltage control, cell characteristics at the time of poisoning are obtained.

FIG. 3 is a graph showing how, when catalyst poisoning occurs in a fuel cell under the operation of constant current control, cell characteristics at the time of the poisoning are obtained.

FIG. 4 is a graph showing how, when catalyst poisoning occurs in a fuel cell under the operation of constant power control, cell characteristics at the time of the poisoning are obtained.

FIG. 5 is a schematic configuration diagram of a fuel cell power generation system 1 as a first embodiment of the present invention.

FIG. 6 is a structural diagram showing a cell structure of the fuel cell stack 10.

FIG. 7 is a vertical cross-sectional view of a carbon monoxide sensor 30.

FIG. 8 is an initial battery characteristic data D stored in a ROM 84.
It is a graph which shows an example of T0.

9 is a flowchart showing a first half of a control routine executed by a CPU 82 of the electronic control unit 80. FIG.

FIG. 10 is a flowchart showing the latter half of the control routine.

FIG. 11 is a graph showing how the current battery characteristic L1 after catalyst poisoning is obtained.

FIG. 12 is a graph showing an operation permission area E.

FIG. 13 shows a current battery characteristic L1 and a constant power line L2.
7 is a graph showing a state in which an intersection point P with and is added to an operation permission area E.

FIG. 14 is a graph showing how the operating point moves in this example.

FIG. 15 is a graph showing a problem of control in the conventional technique.

FIG. 16 is a graph showing a control problem in the conventional technique.

FIG. 17 is a graph showing effects of the first embodiment.

FIG. 18 is a schematic configuration diagram of a fuel cell system as a second embodiment of the present invention.

FIG. 19 CP of two electronic control units 80, 220
It is a flow chart which shows both control routines performed by U82 and 222.

FIG. 20 is a graph showing how to control the intersection P2 with the line L1 indicated by the current battery characteristic data DT1 in the operation permission region E by controlling the power consumption on the load side.

FIG. 21 is a CP of the electronic control unit in the third embodiment.
7 is a flowchart showing a first half portion of a control routine executed by U.

FIG. 22 is a flowchart showing the latter half of the control routine.

FIG. 23: Poison elimination method menu MN stored in ROM
It is explanatory drawing which shows an example.

FIG. 24 is a graph showing the heat generation amount of a polymer electrolyte fuel cell as an example of control characteristics proportional to the poisoning rate R of the catalyst.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Fuel cell power generation system 10 ... Fuel cell stack 10a ... Anode side gas inlet 10b ... Anode side gas outlet 12 ... Methanol tank 14 ... Water tank 16 ... Reformer 18 ... Fuel gas supply passage 18a ... Branch port 20 ... Fuel gas Exhaust passage 22 ... Inverter 24 ... Secondary battery 26 ... Switching device 30 ... Carbon monoxide sensor 32 ... Output detection circuit 41 ... Electrolyte membrane 42 ... Anode 43 ... Cathode 44 ... Separator 44p ... Flow passage groove 45 ... Separator 45p ... Flow passage Groove 46 ... Current collecting plate 47 ... Current collecting plate 50 ... Electrolyte membrane 52 ... Electrode 54 ... Electrode 56 ... Metal plate 58 ... Metal plate 60 ... Holder 60T ... Detecting terminals 60a, 62a ... Flange 60b, 62b ... Screw 62 ... Holder 62T ... Detection terminal 64 ... Insulating member 64a, 64b ... Screw 66 ... O-ring 68 ... Gas inflow passage 7 ... electric circuit 72 ... voltmeter 74 ... resistor 80 ... fuel cell electronic control unit 82 ... CPU 84 ... ROM 86 ... RAM 88 ... input / output interface 201 ... load system 210 ... load 220 ... load electronic control unit 222 ... CPU 224 ... ROM 226 ... RAM 228 ... Input / output interface DT0 ... Initial battery characteristic data DT1 ... Current battery characteristic data L0 ... Initial battery characteristic L1 ... Current initial battery characteristic MN ... Poison elimination method menu R ... Catalyst poisoning rate

Claims (8)

[Claims] 1. A fuel cell driving apparatus for supplying a reaction gas to an electrode carrying a catalyst to obtain an electromotive force from a chemical reaction of the reaction gas, wherein cell characteristics in an initial state of the fuel cell are stored in advance. Initial battery characteristic storage means, poisoning state detection means for detecting that the catalyst is in a poisoned state, and when the poisoning state detection means detects that the catalyst is in a poisoned state, the fuel From the output signal detecting means for detecting the output signal of the battery, the detected output signal of the fuel cell and the cell characteristics stored by the initial cell characteristic storing means, a ratio according to the poisoning rate of the catalyst is calculated. A fuel cell driving device, comprising: a battery characteristic calculating unit that calculates and obtains a cell characteristic when the fuel cell is in a poisoned state from the ratio and the cell characteristic. 2. The drive device for a fuel cell according to claim 1, further comprising an operation permission area setting means for setting an operation area in which the operation of the fuel cell is permitted under various conditions that limit the operation of the fuel cell. A drive device for the fuel cell provided. 3. The drive device for the fuel cell according to claim 2, wherein the battery characteristic in the poisoning state calculated by the battery characteristic calculating means is set in the operation range set by the operation permission area setting means. And on the limited battery characteristics,
Judgment means for judging whether or not the operating point of control can be moved, and when the judgment means determines that the operation point can be moved, permitting switching of the control system of operation of the fuel cell for realizing the movement of the operating point. A drive device for a fuel cell, comprising: 4. The fuel cell drive device according to claim 3, wherein the battery characteristics in the poisoning state calculated by the battery characteristic calculation means, and the operation range set by the operation permission area setting means. And a driving device for a fuel cell, which comprises an information transmitting means for transmitting the above to an externally provided load controlling means. 5. The fuel cell drive apparatus according to claim 3, wherein when the determination unit determines that the operating point cannot be moved, the safety control controls the operation of the fuel cell to a safe side. A drive device for a fuel cell, comprising: 6. A fuel cell in which a reaction gas is supplied to an electrode carrying a catalyst to obtain an electromotive force from a chemical reaction of the reaction gas, and initial cell characteristics in which the cell characteristics of the fuel cell in an initial state are stored in advance. A storage unit, a poisoning state detection unit that detects that the catalyst is in a poisoned state, and an output of the fuel cell when the poisoning state detection unit detects that the catalyst is in a poisoned state Output signal detection means for detecting a signal, and poisoning rate calculation means for obtaining the poisoning rate of the catalyst from the detected output signal of the fuel cell and the cell characteristics stored by the initial cell characteristic storage means. For detecting catalyst poisoning rate of fuel cell including: 7. A fuel cell system comprising the catalyst poisoning rate detection device for a fuel cell according to claim 6, wherein a load connected to the fuel cell, an auxiliary power supply, and the catalyst poisoning rate detection device. The poisoning rate of the catalyst obtained by
A poisoning rate determination means for determining whether or not it is equal to or more than a predetermined value, and when the poisoning rate determination means determines that the poisoning rate of the catalyst is equal to or more than a predetermined value, A fuel cell system comprising: a poisoning time control means for switching the connection from the fuel cell to the auxiliary power source and executing a poisoning countermeasure for the catalyst. 8. The fuel cell system according to claim 7, wherein the poisoning control means includes a plurality of poisoning elimination means for eliminating poisoning of the catalyst by different methods, and the catalyst poisoning rate. A fuel cell system comprising: a selection drive means for selecting any one of the poisoning elimination means according to the poisoning rate of the catalyst obtained by the detection device and driving the selected poisoning elimination means.